S3900 S4700 dbx DriveRack 260

[boputnam]

1) Anyone having a guess which kind of mic-positions Greg used here?
I would like to try this procedure.
2) Mic close to speaker, or at listening positions?
3) Mic at ear levels, or ground-plane?
4) EXACT same mic-positions all the iterations?

1) Use a quality measurement mic which comes with a response curve measure for it. Earthworks M30 is suitable.
2) I prefer measuring where the listening will be done. In the field, we move the mic(s) around to different locations and average. In the home, I fix one position - the sweet spot - and pin it down. The adjacent positions won't be so different (unless you have really beamy horns).
3) Ibid
4) Ibid.

That would be great information. I will be fine tuning a large DIY three-way speaker in next couple of weeks. I was planning on hauling them outdoors for the initial measurements, but they will be very heavy and if in room is good enough for an engineer the caliber of GT it is more than good enough for a geneticist like myself.

Does anyone know of any useful references for setting up digital XO's and EQ? The D'Appolito text on measuring loudspeakers is about all I can find. I know Self has a giant book on digital crossover design, but I believe it is for the development of the device and not its implementation.

For a three-way that is tri-amped, you must do each band pass separately. DO ONLY ONE SIDE OF THE SYSTEM; copy to the other side. However, first, establish your delay(s). The LF will have the most retarded acoustic arrival time due to the LPF. Measure it's impulse and delay the other band passes to match. Then, tune each band pass separately, striving for flat reproduction of Pink noise. Then, open all band passes and measure the overall reproduction of Pink. Look at the response at the crossover points - if you see summing or subtraction (cancellation) try toggling the phase for one band pass or another. Do not worry your little head about this - just listen to the result and how the response curve looks.

Have fun.

[NWCgrad]

1) Use a quality measurement mic which comes with a response curve measure for it. Earthworks M30 is suitable.
2) I prefer measuring where the listening will be done. In the field, we move the mic(s) around to different locations and average. In the home, I fix one position - the sweet spot - and pin it down. The adjacent positions won't be so different (unless you have really beamy horns).
3) Ibid
4) Ibid.

For a three-way that is tri-amped, you must do each band pass separately. DO ONLY ONE SIDE OF THE SYSTEM; copy to the other side. However, first, establish your delay(s). The LF will have the most retarded acoustic arrival time due to the LPF. Measure it's impulse and delay the other band passes to match. Then, tune each band pass separately, striving for flat reproduction of Pink noise. Then, open all band passes and measure the overall reproduction of Pink. Look at the response at the crossover points - if you see summing or subtraction (cancellation) try toggling the phase for one band pass or another. Do not worry your little head about this - just listen to the result and how the response curve looks.

Have fun.

Thanks for the advice...it should be a fun adventure getting the big beast dialed in for optimal performance. Will start off bi-amped (have passive XO for the HF/MF), next year will upgrade my Peavey VSX to the DBX Venu360 and go fully active. The Peavey has some weird software controlled gain structure that induces noise if used to attenuate the signal for the HF unit. Makes it a poor choice for this application, but it is what it is.